The order matters because each step manufactures the raw material for the next. You cannot advise anyone well until you have been advised, and you cannot vouch for someone publicly with any weight until you have a visible record of following through. Mentoring is a loop rather than a favour: asking, acting, teaching and crediting all feed the same reputation.
Step 1: Ask for something specific and bounded
Nothing else on this list works until one named person has agreed to something specific. A general "will you be my mentor" puts the other person in an awkward spot: the commitment is undefined, so the polite answer is "happy to chat anytime", which means never. Ask for something bounded instead. "You rebuilt the pricing team after the restructure; I'm doing the same at a smaller scale. Could we do 30 minutes a month, and I'll send an agenda the day before?" Name the thing they did that you want to learn, name the cadence, and make declining easy. If they say no, ask them to point you to one other person. Send the calendar invite yourself that same day.
Step 2: Bring one decision, leave with one action
A monthly hour is only worth the prep you put into it. The meeting works when you arrive with one live decision — "should I take the lateral move to support, or wait for a lead role on my current team?" — rather than a status update about how busy you are. Send that question 24 hours ahead so the other person can think; most of the value comes from having to write it in one paragraph, which is also how you discover the decision is easier than it felt. Take notes in their words, not your paraphrase. Pick exactly one action, do it before the next meeting, and open the next meeting with what happened. Mentors disengage when advice vanishes into nothing; a two-line report-back is what keeps them investing.
Step 3: Teach what you learned last year
You already know what made the monthly meeting in step 2 useful, and that is precisely the knowledge a junior person needs from you. Pick someone one to three years behind — close enough that your recent mistakes are still relevant, far enough that you have solved problems they are facing now. Offer the same structure you asked for: 30 minutes a month, agenda sent ahead, one action each time. The most useful thing you can say is rarely your hard-won expertise; it is the thing you figured out eighteen months ago that now feels obvious to you and is completely invisible to them. Ask "what have you tried?" before offering anything. Waiting until you feel senior enough is the standard way this step never happens. The person two years behind you is stuck on the exact problem you solved last spring.
Step 4: Three advisors, three different angles
One mentor is one vantage point. By now you have had several months with the person from step 1, and you can probably name what they cannot see — usually anything outside their own company or function. That gap is what a board fixes. Choose three deliberately contrasting people: one inside your company who knows the internal politics, one in your field at another company who knows the market, and one two levels ahead in a different function who will ask why you want the thing at all. Say plainly what you are assembling and what you want from each; nobody is offended by being one of three. Stagger the months so you are never cramming. The temptation is to recruit people who already agree with you, which produces three yeses and no information.
Step 5: Say the good thing in the room
By the end of this sequence you have a mentee, three advisors and a record of acting on advice, which makes vouching credible rather than performative. Public credit only counts when it is specific and timed: name the person and the contribution in the room where the decision is being made, not in a private note afterward. "The churn analysis Priya ran is why we are cutting onboarding from nine steps to four" changes what people in that room think about Priya. Send the recommendation, make the introduction, nominate the award. Private praise mostly comforts the person giving it. A vague "great work" attached to no decision is not credit. The standard failure is waiting for the annual review cycle to say something you could say on Thursday.
A calendar reminder to check in with your mentor vanishes the moment you dismiss it and leaves no trace of whether anything happened. A printed strip taped inside a desk drawer keeps getting shorter as you cut off each finished step, and it stays there, visibly unfinished, until the last one is gone. The loop closes when someone you mentored starts vouching for somebody else.